Rec, West, Aust Mus, 1984,11 (4): 357·373

The Edge-ground Stone Adze and Modern Counterpartsin the Kimberley Region, Western Australia

K. Akerman* and P. Bindont

Ahstract

Aborigines from the southern Kimberley. Western Australia have provided data on stone and shell

adze bl;ldes found on surface artefact sealters, The methods of h:lfting these arc described andillustrated with examples from Museum collections and the relationship between stone or shell

hladed adzes and contemporary metal bladed adzes is discussed, The distribution of each type in theKimberley RegIon IS presented,

Introduction

Adzes are one of the basic components of the Australian Aboriginal woodworking tool kit.They consist of a straight or slightly curved wooden haft and a blade, the cutting edge ofwhich is perpendicular to, but in the same plane as the centre line of the haft. Evidence forthe distribution of flaked adze blades through time and space in Australia has beensummarized by White and CYConnell (1982: 131), who conclude that these tools originate inlate Pleistocene times and were widely utilized, particularly in the drier interior of thecontinent. Recently, we have examined a series of adzes from the Kimberley Region ofWestern Australia which lies north-west of the continent's arid zone (Figure 1). Here themany kinds of wooden artefacts used prior to white settlement are still manufactured withvarious types of adzes. This paper describes in detail one type of edge-ground stone adze,although other adze forms are mentioned and illustrated.

Chisel, Adze or Knife'?Adzes are usually held in both hands and used with short chopping blows made down­

wards and towards the body of the operator. These blows, relying on the momentumgenerated by the operator's arms and the weight of the tool for their effectiveness, pare thinshavings from the work. Chisels, another component of the Australian woodworking toolkit. are generally mueh lighter tools, and although they perform a similar function to adzes,they are driven through the work by a second tool. Both 'adze' and 'chisel' along with othernames have been applied to the small edge-ground artefact which this paper describes (seeMcCarthy Cl 11/. 1946: 54 for references to earlier names). These names were based on thepresumed function or occasionally the observed use of these implements.

Honorarv Assoeiak. Wc'stern Australian Museum. Franeis Street. Perth. Wc'stern Austr:r11:1 hOOODepartment of Ardlaeologv. Western Australian Museum. Francls Street. Perth, Weskrn Australia hOOO

357

l'he l~dgL'-gr()undStone Adze and Modern Counterparts

• Balgo

Turkey.Creek

N

+

Gregory <tJSail Lake

Scale =1.5,700,000

Kimberley

"t'to

<(c;0,0

°<0° ~il'(' L4~

"iI-9 Q«' A --FI~zll()t( ~4. - -_ - - N.cholsonCROSSING "_ •

~ _ - - - - - HALLS/ CREEK

• Lou.sa• Downs

Chrislm~Creek ._-~

.........

Dampierland

• LagrangeMission

Figure I Map AThe metal adze with a tula type predecessor is found cast of the solid lines however stone tulablades and slugs arc more widely distributed. being found westward to the coast.

Metal adze with tula predecessor cast of _

Tula adzc south of - - __

358

K, Akernldl1 ;lnd P Bindn!l

\e Loulsa

Downs

Gregory \;JSalt Lake

N

+Scale = 1,5,700,000

TurkeYeCreek

Nlcholson

eSalgo

Figure I Map B

Dis(nhu(HlI1Il1 thl' fllL'tal/IiII'/ hLHk'S Illllllw\ c1lls,'lv th;ltlll thl' l'd"L '"nJuml st<lllc' hlades, west IIIthl' dllltl'd lilll'

Fdgc ground storlL' dd/e west of ••••••••••

Ld"c grnulld shl'lI ;ld/L' wlthill _

Metalllill'/ ad/e WL'st Ilt _

359

The Edge-ground Stone Adze and Modern Counterparts

According to older Nyikina, Walmatjeri and Kitja informants (Kimberley tribal groupsstill familiar with these artefacts) the common type of edge-ground flakes were hafted andused as adzes. We have thus chosen 'Edge-ground Stone Adze' as being the most descriptiveand accurate term alluding to both the form and main function of the tool type which wedescribe in this paper. We do not deny its incidental use as knife, chisel, scraper or the like.

Nyikina terms for components of the edge-ground stone adze are:

handle or haft - Kangka, of Beauhinia wood (Lysiphyllum cunninghamii) (Tjikal)stone element - (a) Tjirimi

or(b) Tjintapal- this term is used also for the ground shell adze

smew (a) Tjiliwa -leg sinews of kangaroo or emuor

(b) Nyilingka - tail sinew of kangarooresin from spinifexhummock grass(Triodia sp.) - Limirri

Edge-ground Stone AdzesLarge edge-ground artefacts similar in form to Pacific Island adze blades have been

reported from various parts of Australia (Davidson 1938, McCarthy 1944 and 1976). Gener­ally, adzes from the Pacific Region have asymmetrical blades, the working edge of which isperpendicular to, but offset from the central axis of the stone blade when viewed from theside (Figure 2a). Some of these tools are probably not indigenous and it is likely that theothers are merely variants of edge-ground hatchet blades. There seems to be no evidence forassuming that these objects were hafted in any other manner than is usual for Australianhatchets (Dickson 1976: 259 and 1982). Unlike the edge-ground adzes of the Pacific Region,the diminutive Australian example has its cutting edge in line with the central axis of theartefact (Figure 2b). Also, our ethnographic investigation indicates that in the Kimberley itwas used in a manner similar to that described for other types of adzes seen in the desert byobservers like Spencer and Gillen (1904: 634ft') or Gould 1969: 79).

Several early writers commenting on the stone artefacts of south-eastern Australia,include edge-ground adzes and chisels in a wide 'knife' category along with Bulga knives orskin-scapers. They describe the latter as elongated artefacts with one long straight or slightlycurving edge-ground margin (e.g. McCarthy et al. 1946: 55; McCarthy 1976: 53) (Figure 2c).This is in contrast to the typical edge-round adze where the grinding occurs on one of thenarrower margins (MitcheIl1949: 86). Some examples of Kimberley edge-ground adzes havemore than 50% of their edges modified (Figure 5b), but their highly curved working edgeand small size distinguishes them from Bulga knives, many of which were probably equippedwith a simple gum grip (McCarthy et al. Fig. 299) and not hafted in the way described laterin this paper.

360

K, Akcnnan and I) Blndon

b

c_----------__

5cm

o

a

Figure 2 (a) Two views of an edge-ground stone adze blade from the Paeifie showing eutting edge off­set from eentral axis of tool when viewed form the side. New Zealand. (WAM E.147).

(b) Two views of an Australian edge-gronnd stone adze blade, Christmas Creek Station.(WAM B5524/1)

(e) Skin-seaper, the ground edge on one of the long sides, and presumably used unhaftcd.New South Wales (rc-drawn from Thorpe and MeCarthy, Plate 9).

No Kimberley edge-ground adzes have been seen with battered or pereusston markededges opposite the blade, indicative of use as unhafted wedges or true driven chisels.Examples bearing evidence of this type of attrition have been collected from the KurnellPeninsula (Dickson 1982: 78) and other areas of NSW. (White and O'Connell 1982: 128).Such artefacts arc generally made on elongated pieces of raw material with onc narrow endmargin ground to produce the cutting edge. We suggest that the elongated edge-ground adzefigured here (Figure 3a) was not driven through the work with blows from a hammer or maul,but was used with a two-handed grip in the manner described for adzes. 1n this instance webase our identification on both the lack of battering on the margin opposite to the blade andthe opinion of Aboriginal informants. These men took particular note of the material fromwhich the artefact is manufactured and the fairly acute working edge angle and suggestedthat the artefact was used to shape dishes and spearthrowers made from mar/ay (Erythrinavespertilio) a soft wood. This particular artefact, although uncommon in the Kimberley fitswell within the size range of both edge-ground hatchet blades and edge-ground chisels,however the silicified schist from which the tool is made, while being strong enough to permitits use as an adze is not sufficiently robust to withstand hammer blows without shattering.

361

The Edge-ground Stone Adze and Modern Counterparts

c

a

':"..

::;: ..'

cm

./:.:

;i.::.~:.;.?:: ..b

Figure 3 (a) Two views of an edge-ground stone adze made from a slab of silieified schist and usedunhafted. Palm Springs. Louisa Downs Stn. (UW.A. Anthropology Museum Coil.)

(b) Edge-ground metal adze used unhafted. Battered and folded edges on the end oppositeto the sharpened blade indicate that this artefact has been used as a chisel. Ochred twinegrip. Kimberley. (WAM A14233)

(c) Edge-ground metal adze made from a section of hoop iron. Spinifex resin and twine gripincludes a button. Battered end again indicates use as a chisel. Kimberley. (WAMA14232)

362

..----------------------------- --- -----_.

a

cm

c

oI

10I

cm

20I

Figure 4 (a) Edge-ground shell blade. Collected from a coastal soak north of Beagle Bay Mission Hl 1'17'1. (WAM A23575)

(b) Two views of a hafted shell artefact from Sunday Island ';'Cygnct Bay) collectcd 1'132. C,um seal over string-bound hafting Joint.(WAM9956)

(c) Metal bladed 'nayi' adze. Found with a cache of artefacts in the Oscar Ranges in I'IXO. (U.W.A. Anthropology Museulll CoiL)

The Edge·ground Stone Adze and Modern Counterparts

a b

~•........:•. '.•\ ...

[:11··\ ('

",,,,,,;

""

\\\

ec

d

cm

F'igure 5 (a & b) Edge-ground stone adze bladcs. Christmas Crcck Station. (WAM B5524/3 & B5524/2).

(c) Mctal adzc blade made from scction of hand-shcar bladc.lradcd unhaftcd betwccn twomcn 1/6/RI south of Balgo. (UW.A. Anthropology Muscum CoiL)

(d) Similar mctal shcar bladc tradc itcm. Kimberley. (UW.A. Anthropology Muscum CoiL)

(c) Edgc-ground tlaked point. ground for usc unhaftcd. as a tool to producc narrow groovcsin woodcn artefacts. Old Chcrrabun Station. (WAM B4297)

Edge-ground stone adze blades vary considerably in both size and shape. Large examplesmay be difficult to distinguish from small edge-ground hatchet blades (e.g. McCarthy et al.1946: 55). Some, in terms of their length and width fall into the acceptable range of hatchetheads. However as the adze blades are relatively thin and light in relation to their size, itseems unlikely that they were used as hatchets. future morpho-metrical studies and examin­ation of edge wear will probably clarify the dividing line between large edge-ground adzeblades and small hatchet blades.

Throughout their known geographical range edge-ground adze blades were mostly madeof flakes of tough, medium grained, crystalline, indurated or voleanic rocks, lacking planesof easy fracture and similar to those from which edge-ground hatchets are manufactured(Dickson 1972). The Kimberley examples are made on a variety of materials including ultra­mafics, hornfels, dolerite, silicifled tuffs and occasionally silicified schists. Edge-grindingmay take place directly on suitable otherwise unmodified pebbles or primary flakes, or on

364

a

c

oI

10I

cm

10I

Figure 6 (a) Tula fJaked stone adze hlade hafted on wooden handle with Triodia SI'. resin. Made hv Wirili TJahangari in 19S1 ChrIStmasCreek Station. (lI.W.A. Anthropology Museum Coli.)

(h) Shear hlade section hafted as an adze using copper wire and Triodia sI'. resin. Made hy Tpll1tpnu (Jimmy) Ttamhiltin. Balgo(lI.W.A. Anthropology Museum Coli.)

(c) Metal hladed adze, the hlade forged and ground from a section of metallile. Made hy Wiminvlli, Balgo. I'!X I (UW.A. Anthro­pology Museum Coli.)

(d) Face\md side views of a section of car-spring, with onc cnd forged and filed to conform to the hlade shapc of compositcMade hy David Downs, Fitzroy Crossing 19S2. (WAM A23'i(1)

a

cm

e

b

Figure 7 (a) hlgl'-ground stone adze, made bv PaddY Roe (Nyikina speaking m;lI1) Broollle in 19XI. (U.W.A. Anthropology Museum Coil.)

(b) Metal bladed 'nayi' adze. Blade made from section of hand-shear blade. Wide sphts in the handle have becn uscd with a w(;odcn lire-saw to

produce fire. Collected in 1%1 from Christmas Creck Station. (WAM AUX)())(c) Mctal bladed 'nayi' adze colleetL'd in 19.' I. Blade made from section of hand-shear blade. As with example (b) the blade has been removed

;lIld rc-hafted when thc split widened and released its grip on the blade. Northern part of ('anning Stock Route. (WAM 9741).

flakes that havc been trimmed bv second<lrv percussHlIl fLiking Pieces wlllch have hec'ntrimmed hy pressure fl<lking also occur r<lrelv Kimherlev edgl'-ground ;Id/e bLl(ks h<lvcinsufficient mass for pecking or hammer-dressing blows to havc taken effect a IIII conse­quently this ml'lhod of stone reduction W<lS not used to shape these arteLlcls, Akerman hasnoted the modification of a bif<leially fl<lked point to produce an edge-ground tool (Ilnl): Slfl)although this seems to h<lve been <In unusual else (Figure )e). The use of fl<lkes from theblade of an edge· ground hatchet rl'laining onc ground facet from that <lrtdaet has beennoted. Such flakes could h;lve hroken accidentally from the edge of the Iwtchl'l blade duringheavy chopping or they may result from rc-trimming <I previously damaged hatchet hbdeprior to rc-grinding the edge.

Typical Kimberley eX<lmples of edge-ground adles h;lvc onc of the n<lrrower fLlkemargins bifaeiallv ground to form the tvpiell smooth. sharp and slightly curved CUlling edge.As wc havc mentioned. the ratio of modifiL'd edge to unmodified edge v;lries consider;lblyand rare disc-like ex,unples ground on <Ill edges do oecuI. Very few arc ground ;111 over.

Wc have not yl'l collected sufficientmetriell dat<lto establish a more specific lormal 'type'for this artdact. IlowevcI. the tools <llre;ldy eX<lmlned f<lll into two groups: the morecommon pallern (Figures 2h and)a. h) used for ;Id/i ng. and elongated examples wit h narrowground edges presumably used for engr<lvi ng (A kerman and Bindon: in prep,) (Figurl' )l').

Edge-ground Shell Adzes

Edge-ground segments of a mbust shell were h;tllcd lor use ,IS <Idles ;lIld knivcs InDampierland (Akerman 197)a: IS) (Figure 4). This is an a rea devoid of stone outemps whichcould provide raw material for edge-gmund artdaets. No archaeological exelv<ltion I];IStaken pbeL' in this area so it is not yl'l posslhle to determine whl'lherthese multi-purposl'artefacts replaced an adle with a stone hlade or whether in Llct their use extends Llr hack intimc' From the large numher of shell blades occurring on surface artefact scatters inDampierland wc can <lssume that this composite tool occupied a significant pl<lee in thematerial culture of the tribal groups living there. Ilowever. the artdaet is relativelv easy tomanufacture and the shell clement. heing less durable than stone. probably needed freq uentreplacement. Both these factors may explain the relativcly greater numhers of shell hbdes ascompared to stone hlacles seen on surface sites In their respectivc distribution areas.

Hafting

As we havc mentioned. some of the Llrger examples of L'dge-ground adle hl;ldes (length() cm) may h;lve bl'l'nused without h;tlting hut most arl' too small to have hCl'n uSL,d In thiS

way. Examples as small as I) cm in Width and kss than I cm in length at right angles to thL~

ground edge have been sel'n. The specimens \VC l];Ive l'x<lrlllned lrolll the Kimherley shmvscraper/ad/ing wear at right angles to the bbde (ltgure)a). verv dillcrentto tl];lt kllld of use·wear which IS rough Iv parallel to the sharpenull'd,L'.l'. as is usuallv seen on knilc lLikl:s. Thisis in accord with tilL' ethno,L'.raphic data we I];IVC colkcted lor those tools. l isewe<lr of thekind wc havl' notlCl'd m;l\ not he lound on ;111 thl' soutll-eastl'rn Australian eX;lmples.

%

The Edge·ground Stone Adze and Modern Counterparts

In the Kimberley area, edge-ground adze blades were hafted to a short (14-25 cm) stoutpiece of green Bauhinia wood (Lysiphylllltn cllnninghamii). This wood was chosen becauseit is very elastic and is said to grip the stone element without splitting back during use. A splita few centimeter long is made in the end of this haft. Both sides of the split are then pareddown so that the end section of the haft is roughly oval, the split forming the long axis. Theprepared stone blade was forced into the split without prior fitting of the stone or furtheradjustment to the haft. The joint was then bound tightly with sinew obtained from eitherkangaroo or emu. After drying this lashing was usually but not always covered with Triodiaresin (Figure 7a). In an emergency, the stone could be adequately hafted by jamming it into asplit stick with neither sinew or resin strengthening. The resin seal provides some support butmore importantly is to protect the sinew from chafing whilst in use. Very small stone elementswere hafted directly to a shaft with Triodia resin usually without prior splitting. This haftingmethod is not exclusive to edge-ground stone blades but is also the standard method of haftingflaked stone adzes. Akerman (1975a) describes a similar split stick method for hafting shelladzes.

In the south Kimberley, as elsewhere, the tu la adze flake was hafted on either a specialcurved wooden haft, on a throwing stick, or further south in the desert on the proximal endof the spearthrower (Tindale 1965). Hafting involved embedding the trimmed adze flake ina prepared lump of Triodia resin moulded on to the end of the handle (Figure 6a). Althoughrecorded from the desert margin to the south, adze hafts with a stone mounted in both endshave not been seen close to the Fitzroy River (Davidson and McCarthy 1957: 396).

In south-eastern Australia stone blades and ehips were hafted to short sticks with plantexudates and bindings. Some of these are illustrated by Brough-Smyth (1978, J: 379 & 380figs. 199 & 2(0) and Etheridge (1891: plate vi) amongst others. We illustrate in Figure 8, oneof two similarly hafted stone implements collected from Kununurra in 1974 and said to beused as a knife, a chopper or chisel (C.E. Dortch, pers. comm. 1982). For these functions,the tool could be held in either one or two hands. Known as binbalang by the Miriwungpeoples of the east Kimberley this type of artefact can duplicate some of the functions of theedge-ground stone adze from further west. An unmodified pointed cortical flake struck froma rounded river pebble forms the stone element of the binbalang. This is attached directly tothe wooden handle with Triodia resin. Numerous cortical flakes which could be hafted toproduce this tool can be found in surface scatters of artefacts adjacent to the Ord River.Apart from this occurrence. the distribution of these artefacts has not yet been established.Perhaps some were hafted in split sticks in the method described in this paper although themuseum specimens which we have examined were hafted without the split handle.

Metal Counterparts to the Stone AdzesStone bladed woodworking tools have now been replaced by implements with metal

blades. Near Balgo in 1982 we observed a large trimmed stone flake being used to smooth aspear shaft but this is an uncommon occurrence. Aboriginal woodworkers now utilize a suiteof metal tools which include hatchets. chisels, gouges, rasps and files as well as adzes. Thewooden handles of these metal bladed adzes follow closely the shape of those which had

368

K. Akcrman and P Bmdon

cm

Three views of a hafted stone implemcnt. known as liinhlllllng. Collectcd in Kununurra in 1974.(W!\M A2274K)

stone blades. One metal bladed adze currently used in the Kimberley follows the patternalready described for the tula where the blade is gum hafted to a wooden handle. The metalbladed example is made of a short (c. 8-10 cm) section of hand-shears blade (Figure 5c andd), a portion of a flat file or the metal blade of a carpenter's chisel hafted to a length of curvedwood. Except for a widening of the area to which the blade will be attached, this wooden haftis identical to that which was made to accept a stone adze. The metal is attached to a suitablyprepared platform on the wooden haft by binding with light wire - mostly copper wire fromelectrical power lines or twine or sinew. This binding is usually reinforced with a blob ofTriodia resin (Figure 6b). Occasionally a curved vehicle spring or a similarly shaped strip ofmetal about 45 cm long is made into an adze of this form by simply filing a cutting edge on oneor both ends (Figure 6c). Obviously a blade of such length needs no wooden haft althoughsometimes a hand grip of twine and Triodia resin or a strip of rag is provided. Such adzeswhen sharpened at both ends are the only double-ended adzes found in the Fitzroy Riverarea.

The hafted edge-ground stone adze and its shell-bladed coastal relative have beenreplaced by a section of shear blade or similar piece of metal of about the same thickness. Inthis case, a roughly oval shaped piece of metal between 12 and 15 cm long is mounted in awooden handle about 20 cm long. Hafting is accomplished by driving the metal blade intothe end of the short stout haft of Beauhinia wood (Figure 7b). No other lashings or adhesivesare used to hold the blade in the split formed in the handle during this process, althoughoccasionally a strip of rag may be wrapped around the junction of blade and haft to make amore eomfortable grip when the tool is being used. In order to tighten the hafting, the metalblades are removed and driven into the haft again at right angles to the original split (Figure7c and d). This adze type is known as nayi (Walmatjeri).

369

The Edge-ground Stone Adze and Modern Counterparts

A short metal adze, related to the nayi adze is sometimes seen along the Fitzroy River.Sections of suitable metal about 20 cm long with onc cnd sharpened, were used unhafted asadzes for fashioning softwood artefacts much as the stone example mentioned previously andshown in Figure 3a, was used. Occasionally the metal adzes were provided with a grip of ragor twine (Figure 3b and c) and some with battered ends, were used as true driven chisels(Figure 3b).

Sections of metal suitable for the manufacture of blades for all three adze forms arcincluded in gift exchange between trading partners in Kimberley and adjacent communities.The most important trade routes in the south Kimberley run roughly parallel to the FitzroyRiver with trade goods moving both cast and west (Akerman 1979: 24). The bulk of the shearblade sections appear to originate in the west and south-west Kimberley where for severaldecades the pastoral industry was based on sheep.

DistributionConcentrations of edge-ground stone adze blades arc found in two parts of the continent:

in New South Wales and the adjacent parts of northern Victoria and south-west Queensland,(e.g. McCarthy 1946: 53: Mitchell 1949: H4) and also in the Kimberley Region of WesternAustralia to which wc will pay closer attention later. The edge-ground hatchet is a commonartefact in both these areas, and in fact the known distributions of edge-ground hatchets andedge-ground adzes overlap, but hatchets arc found much more widely.

Flaked adze forms like the tula and elouera exist alongside the edge-ground variety inboth the Kimberley and south-cast Australia (sce for example Davidson and McCarthy1957: 3(6). The adzes which occur in the south-cast and centre of Australia have beenadequately discussed in detail by ethnographers and archaeologists alike (e.g. Spencer andGillen 1904: McCarthy cl al. 1(46). Forms and distribution patterns of naked adzes in thenorth-west of the continent arc less well known, (but sce Dortch (1977) for a review of theiroccurrence in thc cast Kimberley). The distribution maps (Figure 1) show the known extentof stone and metal adzes based on either the tula or edge-ground adze predecessor. In theextreme north-west Kimberley, adzes arc rarely represented in the archaeological record,and there is no reference to them in literature concerning the contact or immediate postcontact material culture suite of this part of the region although other scraping tools arcmentioned. Akerman has pointed out elsewhcre (1979a: 245) the differences in the woodenartefact industries throughout the Kimberley which at least in part explain this omission fromthe stone industries.

This adze less trend in the north continues into the present. Contemporary woodworkerstend to use steel hatchets. rasps, knives and modern chisels (not as adzes) in making shields,coolamons and spearthrowers. On the Dampierland Peninsula, woodworkers who have hadnearly a century pf contact with pearling and boat building industries (mainly luggers anddinghies) also tend to use a European suite of tools rather than adze forms based ontraditional shell prototypes. Tulas, which arc semi-discoidal tlakes with varying amounts ofrc-trimming on the dorsal surface of the distal cnd and designed for hafting, arc foundmainly south of a line extending from Thangoo Station on the west coast to near Yecda

370

K, Akcrman and P, 11lndon

Station on the Fitzroy River, to Ilalls Creek and thence north-easterly into the NorthernTerritory. Occasional specimens arc seen north of this line but these arc rare and there is noevidence that tula adzes occupied an important place in the artefact suite in the central andnorth Kimberley Plateau. Flaked nOlHula adzes occur alongside the tulas. They include theburren adze (semi-discoidal tlake with rc-trimming on the lateral margins of the dorsalsurface) and the somewhat enigmatic 'adze point' (Davidson 19.15: 160-1(2) which lacks thctiny step tlake scars on the ventral surface characteristic of other tlaked adzes throughoutAustralia.

The general locus of the edge-ground stone adze in the Kimberley region appears to beconfined to the Fitzroy River Basin and the coast to the north and south of the rllouth of thisriver, but a few specimens occur outside this area. Our main data comes from that territoryformerly held by the Nyikina speaking peoples (the middle reaches of the Fitzroy River)where the edge-ground adzes arc found in conjunction with both edge-ground hatchets andtu la adze tlakes. On the Dampierland Peninsula, and the coast south of Broome they arcfound in conjunction with edge-ground baler shell adzes and knives (Akerman 1975a: 9.1-104and 1975b: 16-19). Edge-ground adzes and knives arc extremely rare in surface sites in theTjaru tribal area (the open plains to the cast and south-cast of Halls Creek) and furthereastwards. 'fhe more common artefacts on these sites include edge-ground hatchets, tlakedadzes and the remnants of these, called 'slugs' which arc discarded after continuedresharpening renders them too small to be effective.

Condusion

Whilst the few excavations of living sites in the west Kimberley failed to establish anytemporal sequence for the full range of tlaked and edge-ground adze types in the area(B1undeIl1975), Dortch established the presence of adze tlakes in his early Ord valley stoneindustry dated back to about 17,000 years ago (Dortch 1977: 1(9).

Examination of eroding surface sites throughout the region has revealed few edge-groundadzes occurring on sites where tlaked adzes arc common. Where edge-ground adzes arecommon, tlaked adzes arc rare. This suggests that the two adze types occupied a similarplaee in the woodworking tool kit, but over what period of time this situation existed remainsto be seen. However, as the edge-ground chisels are made from rock types also used tomanufacture edge-ground hatchets, the possibility exists that the smaller tools may havebeen made along with the latter implements. Edge-ground hatchets have been recoveredfrom Pleistocene deposits in the Kimberley region and nearby Arnhem Land (Dortch 1977and C'. Schrire 19X2).

The occurrence in adjacent areas of the edge-ground shell adze from the DampierlandPeninsula and the edge-ground stone adze of the lower Fitzroy Basin, which were bothhafted in a similar fashion, suggests a close relationship between the two forms. Figure Ibshows very clearly the overlap between these two types and the metal bladed nayi adze. Weconclude that these congruent distributions are more than coincidental and that they reflectthe formal similarity between this metal bladed adze and those with edge-ground stone orshell blades.

371

The Edge-ground Stone Adze and Modern Counterparts

Our observation of the present geographical occurrence of the metal bladed adze whichreplaced the stick hafted tula adze, indicates that the distribution has narrowed from that ofits stone bladed counterpart. This is partly due to the centralisation of population, initially tostations and later to larger settlements, and partly to the heavy influence of Europeanwoodworking tools and skills. We believe that the distribution pattern that emerges demon­strates a conservatism in material culture which allowed the adoption and continuation oftwo distinct adze forms in the southern Kimberley.

Acknowledgements

We thank leff Flenniken, Washington State University, Mance Lofgren and Moya Smith ofthe Western Australian Museum and Peter White of Sydney University for commenting onan earlier draft of this paper. Charles Dortch of the Western Australian Museum deservesspecial thanks for providing data, encouragement and many useful comments.

Bibliography

Akerman, K. (1975a). Baler shell implements from North West Australia. Mankind 10: 16-19.Akerman, K. (l956b). Aboriginal camp sites on the western coast of Dampierland, Western Australia. Gcms.

Paper in Anthropology 4: St. Lucia: Anthropology Mus., Uni. of Qld.Akerman, K. (1979a). Material culture and trade in the Kimberleys today. In: Aborigines of the West: Their Past

and Present (Eds. R.M. & CH. Berndt) (University of Western Australia Press: WA.).Akerman, K. (1979b). An Unusual Stone Chisel from Cherrabun Station, West Kimberleys, WA. Artefact 4: 81-83.Alien, H. (1972). Where the crowflies backwards: man and land in the Darling Basin. Unpubl. Ph.D. thesis. A.N.U.

Canberra.Blundell, UJ. (1975). Aboriginal Adaptation in Northwest Australia. Unpubl. Ph.D. Thesis. Univ. of Wisconsin:

U.S.A.Brough Smyth, R. (1876). The Aborigines of Victoria. 2 vols. Melbourne: John Currey, Q'Neil (facsimile 1972).Davidson, D.S. (1935). Archaeological Problems of Northern Australia. Jour. Roy. Anthrop. Inst. of Gt. Brit. and

I. Vol. 65.Davidson, D.S. (1938). Stone Axes of Western Australia. American Anthropologist 40:38-48.Davidson, D.S. and McCarthy, ED. (1957). Thc distribution and chronology of some important types of stone

implements in Western Australia. Anthropos. 52: 390-458.Dickson, EP. (1972). Ground edge axes. Mankind 8: 206-211.Dickson, EP. (1976). Design and typology of stone hatchets. Mankind 10: 259-263.Dickson, EP. (1982). Australian Stone Hatchets, a Study in Design & Dynamics. Academic Press: Sydney.Dortch, CD. (1977). Early & late stone industrial phases in Western Australia. In: Stone Tools as Cultural Markers:

change, evolution and complexity (Ed. RVS. Wright) (Aust. Inst. Ab. Studies: Canberra).Etheridge, RJun. (1891). Notes on Australian Aboriginal Stone Weapons and Implements. Proc. Linn. Soc.

N.SW 6: 31-43.Gould, R.A. (1967). Yiwara: Foragers of the Australian Desert. Col1ins: London-Sydney.Kenyon, A.S. and Stirling, D.L. (1900). Australian Aboriginal stone implements. A suggested classification. Proc.

Roy. Soc. Vic. 13: 191-2tJO.McCarthy, ED. (1944). Adzes and adze-like implements from eastern Australia. Rec. West. Aust. Mus. 21: 267-271.

372

K_Akerman and P_ Bindon

McCarthy, FD_ (1976)_ Australian Ahoriginal Stone Implemenls_ Austrahan Museum Trust: Sydncy_MeCarthy, F D_. BrarnelL Ic. and Noone, HVv. (1946). The Stone Implements of Australia_ AIemoir '/ (Austrahan

Museum: Sydney)_Mitehell. S_ R. (1949)_ Stone-Age ('rafismeIL '!ilit: Melbourne.Noone, H Vv. (1943)_ Some Aboriginal stone implements of Western Australia. Rec. Sowh Aust. Ml/s. 7: 271-2HO.Sehrire, C (19H2). The Alligator River prehistory and ecology in western Arnhem Land. Terra Al/stralis 7.Spencer, B. and C;illen, I--J. (1904). The Northern Trihes of Central Australia. Maemillan & Co.: London.Thorpe. WA. and MeCarthy, ED. (1933). Ethnological Notes, No. 5. Rec. Al/st. Mm. 19: 23-27.

---------------------------------~-

Received 31 March 1983 Accepted 18 October 1983

373

Published 22 October 1984